Why Making Learning Easier May Not Be the Answer

Can students still learn deeply if AI does much of the cognitive work for them?

This question has surfaced repeatedly in my conversations with higher education teachers over the past year. Many are wondering how learning changes when students can ask AI to explain concepts, generate ideas, write essays, analyse data, and complete assignments in seconds.

The discussion often focuses on AI itself—its capabilities, limitations, and implications for assessment. Yet I believe there is a more fundamental question we should ask first:

What actually makes learning happen?

Rather than debating AI, perhaps we should return to the foundations of learning. What do we know about how people develop understanding, judgement, and expertise? And what does that mean for the way we design learning in higher education?

These questions formed the basis of my presentation at the NOFOMA 2026 Educators' Day in Bergen in June 2026. Although I use supply chain management as the context, the ideas are relevant far beyond a single discipline.

Supply chain management is all about flow. It is about ensuring the smooth movement of materials, information, decisions, and processes across increasingly complex networks. Delays, bottlenecks, and unnecessary handovers are problems to be solved. Success is often measured by how seamlessly things move from one point to the next.

The same principle has shaped many of the world's most successful digital services. Whether we shop online, stream music, watch films, or manage our finances, the best user experiences are often those that require the least effort. One click replaces many, recommendations replace searching and automation replaces routine decisions. The underlying design principle is remarkably consistent: reduce friction, reduce effort, increase usability. In a way, this is not very surprising.

Our brain represents only about two percent of our body weight, yet it consumes roughly twenty percent of our energy. To use that energy efficiently, it constantly seeks shortcuts, familiar patterns, and ways of reducing cognitive effort. Research on cognitive fluency, particularly by Norbert Schwarz and colleagues, has shown that information that is easier to process simply feels better. Ease is not only practical—it is psychologically rewarding.

Artificial intelligence is the latest expression of this long-standing trend. If previous technologies reduced physical work or simplified access to information, AI goes one step further by reducing cognitive effort itself. It can summarise, explain, generate ideas, analyse information, and perform many of the tasks that were previously regarded as evidence of learning.

It is hardly surprising that students have embraced these tools with enthusiasm.

But this raises an important question.

If effort is part of the mechanism through which learning occurs, what happens when effort becomes optional?

Before attempting to answer that question, let's look at something rather curious. Although we generally appreciate smooth, effortless experiences, we also accept—and sometimes even expect—small moments of deliberate friction.

Think about online banking. Making a payment takes only seconds, but before the transaction is completed, we are asked to confirm what we are about to do. The extra step is not there to annoy us, but to protect us. Another sample: speed bumps. Few people enjoy slowing down, yet few would argue that roads near schools should be completely free of them. The temporary inconvenience serves a much more important purpose, that of protecting and ensuring safety.

These examples illustrate an important principle: friction is not inherently good or bad.

Sometimes it is waste that should be eliminated, whereas some other times it is deliberately designed because it improves decisions, increases safety, prevents costly mistakes, or encourages reflection.

Learning belongs much closer to the latter category. Learning is not simply the transfer of information from teacher to student, but rather a process of building understanding, developing judgement, and learning to apply knowledge in new situations. Those processes require learners to think, retrieve, compare, question, make decisions, and occasionally struggle.

In other words, some of the very things that make learning feel more demanding may also be the things that make it more effective. Research in cognitive psychology and the learning sciences has repeatedly pointed in this direction.

Learning Is More Than Fluency

Research in cognitive psychology and the learning sciences has consistently challenged one common assumption: learning does not necessarily improve when it feels easier. Instead, several decades of research point towards a remarkably consistent conclusion: the mental effort required during learning often contributes to learning itself.

Here are four insights that I believe are particularly relevant as we rethink higher education in the age of AI.

1. Ease can be misleading

Robert and Elizabeth Bjork introduced the concept of desirable difficulties to describe learning conditions that may feel more demanding in the moment but lead to stronger long-term learning.

The point is not to make learning unnecessarily difficult. Rather, some forms of challenge encourage learners to pay closer attention, retrieve knowledge, compare alternatives, recognise patterns, and adapt their thinking. These are precisely the mental processes through which understanding develops.

Interestingly, these learning experiences often feel less successful while they are happening. The effort can create the impression that little learning is taking place, even though the opposite is true.

2. Good performance is not always good learning

One of the most important distinctions in learning research is the difference between performance and learning.

Students may appear to perform well during a lecture or guided exercise. They may follow the explanation, answer questions correctly, and complete the activity successfully. Yet weeks later they may struggle to apply the same knowledge in a different context.

As Nicholas Soderstrom and Robert Bjork point out, fluent performance during instruction is often a poor predictor of durable learning.

Learning should be measured not by what students can do while being guided, but by what they can do independently afterwards.

3. We remember what we retrieve

Perhaps one of the most robust findings in educational psychology is the value of retrieval. Research by Henry Roediger and Jeffrey Karpicke shows that trying to recall knowledge strengthens learning more effectively than simply reviewing the same material repeatedly.

In other words, students learn more by using knowledge than by merely being exposed to it. Retrieval requires effort, and that effort is part of what makes learning stick.

4. Understanding grows through experience

Knowledge becomes considerably more useful when learners have opportunities to apply it.

David Kolb's work on experiential learning reminds us that deep understanding develops through a continuous cycle of experience, reflection, conceptual understanding, and renewed action.

It is one thing to understand supply chain concepts in theory, and another thing to experience how a purchasing decision influences production, inventory, customer deliveries, and ultimately financial performance. As a consequence, learning moves beyond remembering facts towards developing judgement.

Taken together, these insights suggest a simple but important conclusion: the goal of education should not be to eliminate all cognitive effort, but to ensure that effort is invested where learning actually takes place.

Why Supply Chain Management Is Different

This brings us to another important question: if learning requires students to think, retrieve, make decisions, and experience consequences, how do we create learning environments that genuinely support those processes?

For supply chain management, this is not a trivial challenge.

Unlike many subjects, supply chain management is fundamentally about interconnected systems rather than isolated concepts. Purchasing decisions influence production that in turn influences inventory. Inventory influences deliveries that influence customer satisfaction, cash flow, and ultimately financial performance.

These relationships are dynamic, often non-linear, and full of trade-offs. Improving one part of the system may unintentionally weaken another. This complexity is difficult to appreciate if it remains confined to lectures, textbooks, or slide decks.

While students may understand individual concepts perfectly well they may still struggle to see how those concepts interact when decisions have to be made in real time. This is where simulation-based learning offers interesting insights and practical solutions.

In simulation-based learning students cannot simply describe supply chain processes; the simulation game mandates them to manage them. Learners must analyse information, discuss alternatives, make decisions under time pressure, coordinate with others, respond to changing conditions, and experience the consequences of their choices. The challenge is not simply remembering concepts, but learning to use them.

In this sense, simulation games create exactly the kinds of productive effort that learning research encourages. They require retrieval, application, collaboration, judgement, adaptation, and reflection—not as separate classroom activities, but as an integrated part of solving authentic problems.

RealGame is one example of this approach. Working in small teams, students manage a real-time operated simulated company and its supply chain. They make operational decisions, interpret business intelligence, respond to changing market conditions, and continuously balance competing objectives. The simulation does not simply illustrate supply chain complexity - it allows students to experience it.

Perhaps that is one of the greatest strengths of simulation-based learning: inviting students to experience what it is to manage dynamic processes in real time.

What Does This Mean in the Age of AI?

While the discussion about AI in higher education often begins with technology, I think a more fruitful starting point is learning and how it can be fostered.

Students already use AI – the genie cannot be put back in the bottle. Nor is the question whether AI has educational value; it clearly does. Used thoughtfully, it has enormous potential to support learning.

The more important question is different:

Which forms of effort should AI remove, and which forms of effort should remain because they are essential for learning?

This distinction becomes particularly important when we consider what universities are trying to achieve. At one level, students need to acquire knowledge, and AI can often help them do this more efficiently. At a second level, students need to learn how to apply that knowledge. Here, AI can become a valuable learning partner by offering explanations, examples, and alternative perspectives.

But there is a third level that deserves even greater attention: higher education is also about developing professional judgement. Professionals rarely encounter textbook problems with one correct answer. Instead, they interpret incomplete information, balance competing objectives, collaborate with others, justify decisions, and act under uncertainty. These capabilities develop through hands-on experience. They cannot simply be downloaded.

This is one reason why simulation-based learning may become even more valuable in the age of AI.

Simulation games require learners to engage in the very activities that turn knowledge into actionable skills and expertise – expertise that will be demonstrated concrete through execution, in context, and often under time pressure.

In other words, simulation creates opportunities for students to develop not only knowledge, but also professional judgement.

Perhaps that is where one of the greatest opportunities for higher education now lies: returning to the fundaments of what learning is and how it takes place, understanding which forms of effort are so essential to learning that we should deliberately sustain them. The goal, after all, is not a frictionless learning experience, but a well-designed learning experience. We should remove the friction that distracts from learning while carefully preserving the effort that develops understanding, judgement, collaboration, and professional expertise.

Making learning easier is not always the answer; designing learning better may be.

As AI becomes increasingly capable of thinking for our students, what kinds of learning experiences will still require students to think for themselves?

References and further reading:

Bjork, E. L., & Bjork, R. A. (2011). Making things hard on yourself, but in a good way: Creating desirable difficulties to enhance learning. 

Soderstrom, N. C., & Bjork, R. A. (2015). Learning versus performance: An integrative review. Perspectives on Psychological Science, 10(2), 176–199.

Roediger, H. L., III, & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255.

Kolb, D. A. (1984). Experiential Learning: Experience as the Source of Learning and Development. Englewood Cliffs, NJ: Prentice-Hall.

Schwarz, N. (2004). Metacognitive experiences in consumer judgment and decision making. Journal of Consumer Psychology, 14(4), 332–348.

Karpicke, J. D., & Blunt, J. R. (2011). Retrieval practice produces more learning than elaborative studying with concept mapping. Science, 331(6018), 772–775.